Civic engagement has been associated with diverse health-related phenomena such as self-reported health, mortality, suicide, and teenage pregnancy. Several causal mechanisms have been proposed for this, but they have been almost exclusively focused on environmental rather than biological or genetic causes, and they have failed to explain a sizable portion of the variance in civic behavior. Thus, we recently investigated the genetic basis of civic behavior of monozygotic and dizygotic twins in the Southern California Twin Register and found that 60-70% of the variation in voter turnout could be explained by genotypic variation. As a result, we hypothesize that heritability may play a strong role in the general predisposition to participate in civic activities, which may help to explain its association with health outcomes. Here, we propose to use the National Study of Adolescent Health ("AddHealth") to explore the causal mechanisms underlying the relationship between civic engagement and health with an emphasis on understanding the role of social networks and their genetic precursors. AddHealth has a base sample size of 90,118, and collected three waves of data between 1994 and 2002. We have four specific aims. First, we will study phenotypic variation in measures of civic engagement in twins (in a sub-sample of 2,030 individuals) to assess the extent to which genes play a role in voting, civic activities, and community involvement. We will also study how civic engagement is affected by genes implicated in social behavior. Second, since structural differences in local social networks are associated with variation in civic engagement, we hypothesize that there may be a link between genotypic variation and social network position. Friendship nominations in AddHealth permit us to recreate the school-level social network and to calculate centrality, clustering, and other network statistics for each respondent. We can then use these measures to study the genetic precursors of social network position. Third, given the strong relationship between social support and health, we hypothesize that at least some of the association between civic engagement and health can be explained by social network position. By integrating measures of social structure into models that show a relationship between health and civic engagement, we will be able to learn if civic engagement is merely a proxy for social networks or if it has an independent association with health behaviors like smoking, drinking, and obesity. Fourth, we will identify the extent to which civic engagement and social structure intermediate the relationship between genes and health behaviors. Genes clearly play a role in social behavior;thus, we hypothesize that at least some of the relationship between genes and health is intermediated by genetic variation in the formation of social ties and the propensity to participate in community activities. Our work has implications for health behaviors like obesity, smoking, and drinking, which are major causes of morbidity and mortality in the US. Gaining a better understanding of the impact of civic behaviors and social networks on health behaviors is therefore of substantial clinical and policy significance.